Why delivering on the AI promise hinges on data and trust

For AI to meet expectations, firms must focus on cleaning their data, eliminating silos, and ensuring a holistic view of customers

2024 was predicted to be the year that businesses harness generative AI to deliver greater productivity, strengthen customer relationships and drive higher margins.

Yet, the reality tells a very different story. Recent studies show that, in most cases, generative AI adoption among corporations has not gone far beyond pilots and technical proof of concepts (PoC) on limited use cases. Many businesses don’t know how AI can add value, but feel pressured to keep up with their competitors and ride the latest technology trend. As a result, many businesses appear to underestimate what they need to use generative AI effectively.

This begs the question: What do businesses really need to deliver on the AI promise? How can they harness generative AI to elevate their business and build competitive advantage?

A key factor in AI adoption is trust. According to our recent AI Trust Quotient research, 4 in 10 Singapore workers don’t trust the data used to train AI systems, and 7 in 10 say it lacks the information needed to be useful. This AI trust gap is hindering adoption in the workplace, preventing businesses from putting AI to full use.

This lack of trust is not unfounded. There are many ways in which AI can go wrong, especially when it is implemented quickly and without the appropriate guardrails. For instance, AI may unintentionally expose private customer data, infringe copyright, or even violate data regulatory compliance requirements – this limits its usefulness and poses negative implications for the business.

Based on research from Forrester, 92% of Singapore respondents believe trust to be important – or even critical – when partnering with an AI vendor. To mitigate risks and build trust in AI, businesses should use AI tools with in-built features like data masking, toxicity detection and zero data retention - to ensure that customer data is kept confidential and that AI results generated are safe and accurate.

While we’re taking quantum leaps forward with AI development, it remains difficult to eliminate every instance of inaccuracy, toxicity, or misinformation. As we look towards an autonomous future and embrace AI-powered tools like AI agents, it is important that we lead with trust. AI tools with powerful system-wide controls that empower humans to focus on the highest-risk, highest-judgement decisions – and delegate the rest – will go a long way in ensuring quality, accurate, and trusted AI outputs.

AI is only as good as the data powering it. Having a strong data foundation ensures that AI can produce contextual, meaningful, and accurate outputs that are useful to workers. Otherwise, the AI output generated can be fairly generic and requires editing and personalisation before it can be used.

What holds businesses back from tapping into generative AI more effectively is the fact that much of their data exists in silos and is disconnected from the customer layer – an organisation has an average of 991 business applications, of which, over 72% are disconnected. This is a challenge for many businesses in the region. In Singapore for example, while the majority of business leaders acknowledge the importance of a robust data strategy, only 30% claim to have one implemented across their business.

For AI investments to meet expectations, businesses must focus on cleaning their data, eliminating silos, and ensuring a holistic view of customer data.

In this new AI era -- where customer data and metadata are considered the new gold -- CRM is the goldmine. CRM plus a data platform that unifies all of the company’s structured and unstructured data, brings the power of AI into the hands of business users. This makes AI actionable by the front office and unlocks the value of both generative and predictive AI, bridging the value gap.

Today, AI-powered CRMs are revolutionising how businesses use AI to enhance productivity and personalisation, directly in the flow of work. Done right, it can help organisations build initial AI use cases and drive AI adoption, acting as a springboard for enterprise-wide AI transformation.

For instance, sales teams can use natural language prompts to summarise records and accelerate deal closure, or generate customised communications for personalised customer engagements. Service teams can streamline case resolution and boost customer satisfaction by surfacing relevant answers and data from disparate systems.

Take Converge ICT as an example. The leading fibre broadband provider in the Philippines is well underway in establishing one of the first generative AI contact centres in the Philippines, using predictive and generative AI through Salesforce’s Einstein 1 platforms. With AI, Converge ICT is increasing operational efficiency and enhancing customer digital experiences.

There remains real potential in harnessing AI to drive efficiency, productivity, and personalisation to transform businesses. However, businesses must put in the groundwork, and develop an AI strategy built on a strong foundation of data and trust. Being use case-driven will allow businesses to roll out features correctly, with solid adoption and measurable value, to deliver on the AI promise.